Influence of Prior Nicotine and Alcohol Use on Functional Outcome in Patients after Intracerebral Hemorrhage

Abstract

The influence of prior nicotine or alcohol use (legal drug use [LDU]) on outcome measures after intracerebral hemorrhage (ICH) is insufficiently established. We investigated drug-specific associations with (1) neuroradiologic and clinical parameters and (2) functional long-term outcome after ICH. This observational cohort study analyzed consecutive spontaneous patients with ICH (n = 554) from our prospective institutional registry over a 5-year study period (January 2010 to December 2014). We compared no-LDU patients with LDU patients, and patients using only nicotine, only alcohol, or both. To account for baseline imbalances, we reanalyzed cohorts after propensity score matching. Prevalence of prior LDU was 197 of 554 (35.6%), comprising 94 of 554 (17.0%) with only nicotine use, 33 of 554 (6.0%) with only alcohol use, and 70 of 554 (12.6%) with alcohol and nicotine use. LDU patients were younger (65 [56-73] versus 75 [67-82], P <.01), less often female (n = 61 of 197 [31.0%] versus n = 188 of 357 [52.7%], P <.01), had more often prior myocardial infarction (n = 29 of 197 [14.7%] versus n = 24 of 357 [6.7%], P <.01), and in-hospital complications (sepsis or systemic inflammatory response syndrome: n = 95 of 197 [48.2%] versus n = 98 of 357 [27.5%], P <.01; pneumonia: n = 89 of 197 [45.2%] versus n = 110 of 357 [30.8%], P <.01). Except for an increased risk of pneumonia (odds ratio 2.22, confidence interval [1.04-4.75], P = .04) in patients using both nicotine and alcohol, we detected no significant differences upon reanalysis after propensity score matching of neuroradiologic or clinical parameters, complications, or long-term outcome between patients with and without LDU (mortality: n = 48 of 150 [32.0%] versus n = 45 of 150 [30.0%], P = .71; favorable outcome [modified Rankin Scale 0-3]: n = 56 of 150 [37.3%] versus n = 53 of 150 [35.3%], P = .72). Prior nicotine use, alcohol use, and their combination were associated with significant differences in baseline characteristics. However, adjusting for unevenly balanced baseline parameters revealed no differences in functional long-term outcome after ICH.